fix: db compaction run (#1156)

These fixes were taken from https://github.com/arneman/EOS/tree/refactor/economic-objective.

1. Fix compaction job registration bug in eos.py:
   - compact_eos_database was registered with save_eos_database function
   - Now correctly calls compact_eos_database()
   - Root cause: compaction/vacuum never ran, allowing records to grow unbounded

2. Optimize db_iterate_records() from O(n) to O(log n):
   - Previous: linear scan from index 0 for every key_to_array() call
   - Now: uses bisect_left on _db_sorted_timestamps to skip to start position
   - Critical for large datasets: 9400 measurement records → 94s/call overhead → 5 calls/optimization

3. Reduce compaction_interval_sec default from 604800s (7 days) to 3600s (1 hour):
   - With 7-day interval: 65000 records accumulate before first cleanup
   - Bridge pushes ~1 record/9s (grid_export_emr)
   - Hourly compaction + 2h data window → stable ~950 records

Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com>

src/akkudoktoreos/core/database.py

@@ -109,7 +109,7 @@ class DatabaseCommonSettings(SettingsBaseModel):
     )
 
     compaction_interval_sec: Optional[int] = Field(
-        default=7 * 24 * 3600,  # weekly
+        default=3600,  # hourly
         ge=0,
         json_schema_extra={
             "description": (
@@ -117,7 +117,7 @@ class DatabaseCommonSettings(SettingsBaseModel):
                 "Compaction downsamples old records to reduce storage while retaining "
                 "coverage. Set to None to disable automatic compaction."
             ),
-            "examples": [604800],  # 1 week
+            "examples": [3600],  # 1 hour
         },
     )
 

src/akkudoktoreos/core/databaseabc.py

@@ -1603,7 +1603,14 @@ class DatabaseRecordProtocolMixin(
             end_timestamp=end_timestamp,
         )
 
-        for record in self.records:
+        # Use bisect to find the first record at or after start_timestamp.
+        # _db_sorted_timestamps and self.records are always kept in parallel
+        # (same insertion index), so bisect_left gives the correct slice start.
+        start_idx = 0
+        if start_timestamp and not isinstance(start_timestamp, _DatabaseTimestampUnbound):
+            start_idx = bisect.bisect_left(self._db_sorted_timestamps, start_timestamp)
+
+        for record in self.records[start_idx:]:
             record_date_time_timestamp = DatabaseTimestamp.from_datetime(record.date_time)
 
             if start_timestamp and record_date_time_timestamp < start_timestamp:
@@ -1702,7 +1709,7 @@ class DatabaseRecordProtocolMixin(
                     f"Vacuum requested for database '{self.db_namespace()}' but keep limit is infinite."
                 )
                 return 0
-            keep_hours = keep_duration.hours
+            keep_hours = int(keep_duration.total_seconds() // 3600)
 
         if keep_hours is not None:
             _, db_max = await self.db_timestamp_range()

src/akkudoktoreos/server/eos.py

@@ -184,7 +184,9 @@ async def lifespan(app: FastAPI) -> AsyncGenerator[None, None]:
         "save_eos_database", save_eos_database, interval_attr="database/autosave_interval_sec"
     )
     manager.register(
-        "compact_eos_database", save_eos_database, interval_attr="database/compaction_interval_sec"
+        "compact_eos_database",
+        compact_eos_database,
+        interval_attr="database/compaction_interval_sec",
     )
     manager.register("manage_energy", ems_manage_energy, interval_attr="ems/interval")